ci: route build IO through Dev Drives - #31357
Conversation
2a2c345 to
251e09b
Compare
7163e06 to
1c3f5f3
Compare
251e09b to
088856e
Compare
1c3f5f3 to
c33c58b
Compare
088856e to
1bb73d7
Compare
c33c58b to
b651571
Compare
b651571 to
9a9175f
Compare
## Why Prepare CI jobs for a later build-output relocation without changing where they write today. ## What - Export `CARGO_TARGET_DIR` from `setup-ci` at the existing `codex-rs/target` path. - Route nextest, release, artifact, and signing paths through `CARGO_TARGET_DIR`. - Require V8 staging callers to pass an explicit target directory while preserving the existing upstream path. ## Manual validation - Ran `just test-github-scripts`. - Parsed GitHub Actions YAML with `yq`. ## Stack - [#31332](#31332) — parameterize Cargo target paths - [#31356](#31356) — Windows 2025 runner bump - [#31357](#31357) — Dev Drive I/O routing
5daeb21 to
97b4767
Compare
9a9175f to
f245102
Compare
## Why The V8 Windows source jobs need a runner image that can support the later Dev Drive setup. ## What Move the two V8 Windows source-build matrix entries from `windows-2022` to `windows-2025`. Namespace the canary source-build cache by runner image so it cannot restore Windows 2022 native outputs. ## Manual validation - Ran `just fmt`. - Ran `just test-github-scripts` (33 tests). - Parsed GitHub Actions YAML with `yq`. - Ran `git diff --check`. ## Stack - [#31332](#31332) — parameterize Cargo target paths - [#31356](#31356) — Windows 2025 runner bump - [#31357](#31357) — Dev Drive I/O routing
f245102 to
3b99045
Compare
02ff4e3 to
c3d96ee
Compare
|
@codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: c3d96ee6d3
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
@codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: dedbddbecd
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Why
Windows Cargo and Bazel jobs spend significant time in filesystem-heavy build and cache directories. Route those directories through one CI build root so Windows can use its Dev Drive and Unix can use a stable cache root.
What
setup-cidefineCI_BUILD_ROOT,CARGO_TARGET_DIR, Bazel cache/output paths, and temp paths.C:.setup-bazelso its explicitoutput_basedoes not defeat Dev Drive routing.Benchmark results
One-off cold-cache WPR/ETW traces show the explicit Bazel output-base routing removes the dominant
C:traffic:C:\_bazelC:trafficFor a cold x64 V8 source build, the retained build-tail sample showed
D:\cargo-targetat ~1.29 GiB while measuredC:roots totaled ~0.45 GiB (C:\Users~0.33 GiB,C:\Program Files~0.06 GiB,C:\Windows~0.03 GiB). The full cold build took 2h20m36s.The Bazel timing improvement is directional because both refreshed shards failed tests. The V8 trace is a bounded build-tail sample, not the full build. All final samples had zero lost ETW events; VHDX traffic was excluded from the optimization ranking.
Runs: baseline Bazel, updated Bazel, V8 build tail.
Manual validation
just fmt.just test-github-scripts(35 tests).yq.git diff --check.Stack